Steroids and Oocyte Maturation—A New Look at an Old Story

Abstract

Female fertility requires precise regulation of oocyte meiosis. Oocytes are arrested early in the meiotic cycle until just before ovulation, when ovarian factors trigger meiosis, or maturation, to continue. Although much has been learned about the late signaling events that accompany meiosis, until recently less was known about the early actions that initiate maturation. Studies using the well-characterized model of transcription-independent steroid-induced oocyte maturation in Xenopus laevis now show that steroid metabolism, classical steroid receptors, G protein-mediated signaling, and novel G protein-coupled receptors, all may play important roles in regulating meiosis. Furthermore, steroids appear to promote similar events in mammalian oocytes, implying a conserved mechanism of maturation in vertebrates. Interestingly, testosterone is a potent promoter of mammalian oocyte maturation, suggesting that androgen actions in the oocyte might be partially responsible for the polycystic ovarian phenotype and accompanying infertility associated with high androgen states such as polycystic ovarian syndrome or congenital adrenal hyperplasia. A detailed appreciation of the steroid-activated signaling pathways in frog and mammalian oocytes may therefore prove useful in understanding both normal and abnormal ovarian development in humans.

HISTORY LESSON: PROGESTERONE PROMOTES XENOPUS OOCYTE MATURATION IN VITRO

FEMALES FROM NEARLY every species of animal are born with their full complement of oocytes; however, these immature oocytes are arrested in prophase I of meiosis. Just before ovulation, gonadotropins stimulate ovarian follicular development, which in turn promotes oocytes to reenter the meiotic cycle, or mature, until they arrest again in metaphase II of meiosis. These mature oocytes are then competent for ovulation and subsequent fertilization (1–3). To date, much is still unknown about the ovarian signals that both inhibit and promote oocyte maturation, especially in mammals.

The phenomenon of steroid-induced maturation of Xenopus laevis oocytes is one of the best-studied models of meiosis (4, 5). More than 60 yr ago, Rugh (6) induced ovulation in amphibians by treating females with pituitary extracts. Shortly afterward, others found that ovulation could be induced by treating female frogs with ovarian tissue that had been pretreated with pituitary extracts (7–9), suggesting that some component of the ovarian tissue, and not a pituitary factor, induced ovulation. In 1968, Dennis Smith reported that follicular cells from frog ovarian tissue produced a factor(s) that could induce ovulation (10). He proposed that progesterone was the ovulation-inducing factor produced by the follicular cells, as submicromolar concentrations of progesterone promoted maturation in vitro.

THE ROLE OF ANDROGENS IN XENOPUS OOCYTE MATURATION

Although progesterone had been the assumed in vivo mediator of Xenopus oocyte maturation for many decades, its physiological importance had not been verified. In fact, many steroids other than progesterone were equally potent promoters of oocyte maturation in vitro (5). This point was particularly important given that progesterone promoted oocyte maturation in other lower vertebrates such as fish, yet progesterone metabolites were actually the primary mediators of maturation in vivo (11, 12).

The physiological role of progesterone in Xenopus ooctye maturation was examined by injecting frogs with human chorionic gonadotropin (hCG) followed by measurement of serum and ovarian steroid levels (13). Surprisingly, progesterone was nearly undetectable, regardless of the dose of hCG or the time after injection that the frogs were examined. Instead, the androgens androstenedione and testosterone, which were equally or more potent promoters of oocyte maturation relative to progesterone, respectively, were abundant. These data suggested that androgens, rather than progesterone, were in fact the dominant mediators of Xenopus oocyte maturation in vivo.

THE OOCYTE IS REQUIRED FOR ANDROGEN PRODUCTION IN XENOPUS LAEVIS

The high androgen and low progesterone production by Xenopus ovaries suggested that either very little progesterone was produced or that progesterone was rapidly metabolized after synthesis. Characterization of the steroid biosynthetic pathway in Xenopus ovaries supported the former alternative, as ovarian androgen production bypassed progesterone and proceeded almost entirely through the Δ5 pathway (14) (Fig. 1). Frog ovaries also rapidly metabolized exogenously added progesterone to testosterone, demonstrating that, although not significantly used in vivo, steroid metabolism via the Δ4 pathway was also possible. Surprisingly, CYP17, the enzyme that converts pregnenolone to dehydroepiandrosterone (DHEA) and progesterone to androstenedione, was expressed exclusively in oocytes, whereas all other steroidogenic enzymes were found in the surrounding follicular cells (14). This suggests that Xenopus oocytes play a critical role in the production of the steroid used to promote their own maturation (Fig. 1). The concept of oocytes regulating their own maturation mirrors studies in mammalian ovaries, where oocytes secrete factors such as growth differentiation factor 9 (GDF9) to promote granulosa cell proliferation and differentiation, which are in turn necessary for normal follicle growth and maturation (15, 16). Depending on the animal, oocytes might therefore utilize different mechanisms to orchestrate proper follicular development and subsequent ovulation.

Fig. 1.

Steroidogenesis in the X. laevis Ovary Steroid production in the Xenopus ovary requires both follicular cells and oocytes. Pregnenolone (PREG) is produced in follicular cells but is a poor substrate for 3β-hydroxysteroid dehydrogenase (3βHSD); thus, virtually no progesterone (PROG) is produced. Instead, pregnenolone is rapidly converted to 17-hydroxypregnenolone (17OHPREG) and then DHEA by CYP17 expressed exclusively in the oocyte (the Δ5 pathway). DHEA, which cannot promote oocyte maturation but is an excellent substrate for 3βHSD, must then reenter surrounding follicular cells for conversion to androstenedione (AD) and testosterone (Test). These two androgens subsequently act back on the oocyte to promote maturation. Because testosterone is more potent and abundant than androstenedione, it is most likely the primary androgen that regulates maturation in vivo. Bold green print and lines represent the dominant physiological pathway for androgen production, although the Δ4 pathway (red) can be used if progesterone is added exogenously.

One final critical concept derived from these metabolism studies is that, because oocytes contain high levels of CYP17 activity, they rapidly metabolize progesterone to androstenedione. As androstenedione and progesterone are equally potent promoters of oocyte maturation, both steroids are likely mediating maturation upon addition of progesterone.

STEROID RECEPTORS IN THE OOCYTE

Identification of receptor(s) involved in steroid-mediated oocyte maturation has been an ongoing challenge for many years. Early evidence indicated that these receptor(s) were near the cell surface, because 1) steroids covalently attached to polymers or BSA, thus unable to enter oocytes, still induced maturation (4, 17); 2) high-affinity steroid binding sites [equilibrium constant (K_d) values in the 10–100 nm range] were detected in oocyte membrane preparations (12, 18–20); and 3) steroid-induced maturation was shown to occur independently of transcription, as it was unaffected by the transcriptional inhibitor actinomycin D (21, 22). Progress toward identifying these membrane-localized steroid receptors was recently accelerated by the discovery that classical nuclear/cytoplasmic steroid receptors associated with plasma membranes mediate nongenomic steroid-induced signaling in other systems. Examples include estrogen-induced activation of MAPK and endothelial nitric oxide synthase in breast and endothelial cells, respectively (23–25), estrogen and androgen-mediated protection from apoptosis in bone cells (26, 27), and testosterone-mediated activation of MAPK and the transcription factor cAMP response element-binding protein (CREB) in Sertoli cells (Walker, W., personal communication).

With these examples in mind, the role of classical steroid receptors in progesterone- and androgen-induced maturation of Xenopus oocytes was explored. Approximately 3 yr ago, two laboratories independently cloned different isoforms of the classical nuclear/cytoplasmic progesterone receptor (PR) from Xenopus oocytes and demonstrated through overexpression and antisense experiments that modulation of intracellular PR levels resulted in small changes in progesterone sensitivity (28, 29). Additionally, a small percentage of PRs (5–10%) were shown to be associated with the oocyte plasma membrane, and coimmunoprecipitation indicated that they might interact with the membrane-signaling molecules MAPK and phosphatidylinositol 3-kinase (30). Together, these studies suggested that the classical PRs might be playing at least a partial role in progesterone-mediated maturation.

In contrast to the studies examining PR, experiments using androgens to promote maturation, thereby eliminating concern over steroid metabolism, have more clearly implicated the classical androgen receptor (AR) as a mediator of androgen-induced events (13, 22). Biochemical and immunohistochemical studies showed that the AR is associated with the plasma membrane of Xenopus oocytes. Further, the AR antagonist flutamide specifically attenuated androgen-mediated signaling and maturation in isolated oocytes, as well as hCG-stimulated maturation of oocytes in intact ovarian follicles. Finally, elimination of endogenous AR by RNA interference specifically reduced androstenedione-mediated maturation in vitro.

Surprisingly, although the AR appears to be an important mediator of testosterone-induced signaling and maturation, some known agonists of AR-mediated transcription are poor promoters of oocyte maturation. For example, R1881, one of the most potent activators of AR-mediated transcription, was unable to promote nongenomic signaling and maturation in oocytes. In fact, R1881 inhibited testosterone-mediated events in oocytes, suggesting that it binds to the AR and acts as a competitive inhibitor of nongenomic testosterone-induced signaling (22). Conversely, androstenediol was a poor promoter of AR-mediated transcription, but a strong mediator of maturation and nongenomic signaling in oocytes (31).

Although classical steroid receptors appear to play some role in steroid-induced maturation of Xenopus oocytes, a new family of potential membrane-associated steroid receptors has been described that may also be important in some animals. The first of these receptors was cloned from oocytes of the spotted seatrout (32, 33). Seatrout oocytes mature in response to progesterone, although the physiological maturation-promoting steroid is the progesterone metabolite 21-trihydroxy-4-pregnen-3-one (20β-S). This novel membrane PR (mPR), is part of a family of proteins that bear structural, but little sequence, homology to the G protein-coupled receptor family. Binding studies using bacterially expressed mPR revealed specific, high affinity binding to progesterone and progesterone metabolites. In addition, mPR expressed in breast cells mediated a pertussis toxin-sensitive decrease in intracellular cAMP in response to progestins, which is consistent with progestin-mediated reductions in fish oocyte cAMP levels, and suggests that mPR may couple to Gα_i. Finally, injection of mPR antisense oligonucleotides into zebrafish oocytes attenuated progestin-induced oocyte maturation, indicating that endogenous mPR may be necessary for oocyte maturation in fish. To date, the role of mPR in mediating oocyte maturation in other animals has yet to be determined.

INTRACELLULAR SIGNALING AND OOCYTE MATURATION—A RELEASE OF INHIBITION MODEL

Oocyte maturation is associated with a myriad of intracellular signals, most of which are conserved across species, suggesting a common mechanism of maturation in vertebrates. Many of these signals accompany maturation regardless of the initiating factor; thus, they may be activated by the maturation process itself rather than directly by steroids.

One critical molecule activated during oocyte maturation in most species is the MOS protein. Mos mRNA is present in oocytes during meiotic arrest; however, little is translated into protein. During maturation, several factors, including Eg2 and cytoplasmic polyadenylation element binding protein (CPEB), alter the polyadenylation of Mos mRNA to allow its translation (34, 35). MOS is a potent activator of MAPK cascade and cyclin-dependent kinase 1, or CDK1 (also called CDC2). Induction of any of these three components promotes activation of the other two, resulting in a powerful positive feedback loop that triggers maturation (36). The necessity of these signals for maturation is controversial, however, as oocytes in which the MAPK cascade has been blocked or MOS expression eliminated still mature under some conditions (37–42). Regardless of their requirement for maturation, the strongly positive effects of all three signals on each other and on maturation suggest that they play significant, but perhaps not exclusive, roles in promoting oocyte maturation.

Of note, MOS, CDK1, and MAPK activation are not detectable until hours after initiation of meiosis. In contrast, cAMP appears to be important earlier in the regulation of maturation. Elevation of intracellular cAMP levels by addition of cAMP analogs or phosphodiesterase inhibitors blocks oocyte maturation in most species, suggesting that cAMP is an important inhibitor of meiosis. Consistent with these findings, some studies have shown that intracellular cAMP levels drop rapidly after initiation of maturation by steroid (43–46), although whether this decrease in intracellular cAMP is either necessary or sufficient to promote oocyte maturation remains controversial (47, 48).

Together, these studies suggest a release-of-inhibition mechanism of oocyte maturation whereby intracellular signals such as cAMP maintain meiotic arrest (Fig. 2). Meiosis may in part be inhibited by G proteins, as constitutive Gβγ and Gα_s signaling appears to hold isolated Xenopus oocytes in meiotic arrest, and both are capable of stimulating adenylyl cyclase to elevate intracellular cAMP (20, 49, 50). This constitutive G protein signaling might be mediated by free endogenous G protein subunits or by an endogenously activated G protein-coupled receptor (possibly a member of the mPR family) (51, 52). Steroids may then trigger maturation by overcoming or blocking these inhibitory signals, thus lowering intracellular cAMP levels and allowing meiosis to progress.

Fig. 2.

Release of Inhibition Model for Oocyte Maturation A, Meiotic arrest is maintained by constitutive signals that elevate cAMP and prevent maturation. In Xenopus, these signals are endogenous to the oocyte, and include Gα_s and Gβγ, which may be signaling independent of or in response to a G protein-coupled receptor (GPCR) (perhaps a member of the mPR family). In mice, these signals may be regulated by an unknown ovarian inhibitor (I) activating a GPCR or some other receptor. Steroids overcome this inhibition, allowing meiosis to progress (B). Two of many possible mechanisms are 1) liganded classical steroid receptor (SR) interrupts GPCR- (left) or free G protein-mediated (right) signaling; 2) competitive binding of steroid directly to the GPCR either decreases Gα_s/Gβγ signaling or, in fish, perhaps activates Gα_i via mPR. Once the inhibitory signals, which include cAMP, are removed or overcome, secondary signals such as Mos, MAPK, and cdc2 are activated. These signals then promote each other in a powerful positive-feedback loop, resulting in germinal vesicle breakdown (GVBD) and maturation.

STEROIDS AND MAMMALIAN OOCYTES

Interestingly, regulation of mammalian oocyte maturation is quite similar to that in Xenopus laevis. Meiotic arrest of mouse oocytes also appears to be regulated by constitutive signals that elevate intracellular cAMP, and the same signaling pathways are activated when mouse oocytes mature (45, 53, 54). In contrast to Xenopus, however, mouse oocytes spontaneously mature when removed from the ovary, suggesting that the primary signal maintaining meiotic arrest of mouse oocytes comes from the ovary rather than being endogenous to the oocyte itself, as with frog oocytes.

Because isolated mouse oocytes spontaneously reenter meiosis in vitro, studies aimed toward identifying factors that promote maturation have been difficult. A meiosis-activating C₂₉ sterol, FF-MAS, which may play a role in enhancing oocyte maturation in vitro, has been described (55). In addition, some studies have shown that micromolar concentrations of testosterone partially inhibit spontaneous maturation under certain conditions (21, 56–58). These studies are all difficult to interpret, however, as they 1) looked primarily at changes in oocytes that were already spontaneously maturating; 2) used oocytes that had been preexposed to high levels of sex steroids (e.g. mice were pretreated with gonadotropins or were postpubertal); and 3) only saw effects with high (micromolar) concentrations of sterol.

To avoid the problems of steroid priming and spontaneous maturation, recent studies were performed using oocytes from prepubertal female mice that were not pretreated with gonadotropins (59). In addition, these oocytes were treated with the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine to maintain meiotic arrest. Interestingly, both testosterone and estradiol were able to overcome this inhibitory signal and promote germinal vesicle breakdown, as well as activation of CDK1 and MAPK, in a dose-dependent fashion (EC₅₀ values in the 100–200 nm range). As in Xenopus, these responses appeared to occur independently of transcription. Further, testosterone and estradiol responses were inhibited by flutamide and ICI 182,780, respectively, indicating that classical steroid receptors might be involved. Finally, as with frog oocytes, the AR transcriptional agonist R1881 did not promote activation of MAPK or maturation in mouse oocytes and, in fact, partially inhibited testosterone-mediated maturation.

Further in vivo studies will be necessary to confirm the physiological importance of steroid-mediated maturation in mice; however, these studies support the “release of inhibition” model oocyte maturation presented in Fig. 2. One could speculate that unknown inhibitory signals (I) within the mouse ovary hold meiosis in prophase I, perhaps by elevating intracellular cAMP through interactions with G protein-coupled receptors. In vitro removal of oocytes from the ovary would eliminate these inhibitory signals, thus resulting in spontaneous maturation. Alternatively, in vivo stimulation by gonadotropins before ovulation would lead to follicle growth and production of sex steroids. Only a few dominant follicles might be expected to produce enough steroid to overcome the inhibitory signals and allow meiosis and subsequent ovulation to progress.

The ability of androgens to signal in mammalian oocytes suggests an intriguing mechanism that might partially explain the polycystic phenotype seen in the ovaries of women with androgen excess from polycystic ovarian syndrome (60), congenital adrenal hyperplasia (61), CYP19 (aromatase) deficiency (62), or exogenous androgen usage (63). Given the importance of cross-talk between oocytes and surrounding cells for normal follicle growth (15, 16), one could speculate that excess or altered nongenomic androgen signaling in oocytes might modify this relationship and contribute to the unregulated follicular growth and lack of dominant follicle production in women with androgen excess. This dysregulation might in turn lead to anovulation and infertility. Whether the ovaries of these patients contain excess mature oocytes is not known; however, the AR antagonist flutamide improves fertility in some women with polycystic ovarian syndrome (60, 64), consistent with androgen signaling through the AR playing a role in their infertility. Further, female mice lacking the AR have reduced fertility (65), suggesting that androgen actions via the AR are important, but not necessary, for normal ovarian physiology.

In summary, steady progress over many decades has revealed a complex set of factors regulating oocyte maturation and ovarian development. By combining lessons learned from the traditional experimental systems in frogs and fish with those more recently from mouse models, it is now possible to begin addressing the role of steroids in human oocyte development. Thus, what began nearly 70 yr ago as an injection of pituitary extracts into frogs has progressed to the potential development of novel modulators of nongenomic steroid-mediated signaling that might precisely regulate oocyte maturation in women with fertility defects.

Acknowledgments

We thank the many researchers who have contributed to the field of oocyte maturation and apologize to those whom we could not reference due to space limitations. We also thank Will Walker for his advice in the preparation of this manuscript.

S.R.H. is a W. W. Caruth, Jr. Scholar in Biomedical Research and is supported by NIH Grant DK59913 and Grant I-1506 from the Welch Foundation.

Abbreviations:

 
• AR,

  Androgen receptor;

 
• CDK1,

  cyclin-dependent kinase 1;

 
• DHEA,

  dehydroepiandrosterone;

 
• hCG,

  human chorionic gonadotropin;

 
• mPR,

  membrane PR;

 
• PR,

  progesterone receptor.